Management of asymptomatic abdominal

ORIGINAL ARTICLE Immediate Open Repair vs in Patients with Small Abdominal Aortic Aneurysms: Survival Differences by Aneurysm Size Abstract Giovanni Filardo, PhD, MPH; Frank A. Lederle, MD; David J. Ballard, MD, MSPH, PhD; Cody Hamilton, PhD; Briget da Graca, JD, MS; Jeph Herrin, PhD; Jessica Harbor, MS; Julie B. VanBuskirk, MPH; Gary R. Johnson, MS; and Janet T. Powell, PhD, MD, FRCPath Objective: To assess whether survival differences exist between patients undergoing immediate open repair vs surveillance with selective repair for 4.0- to 5.4-cm abdominal aortic aneurysms (AAAs) and whether these differences vary by diameter, within sexes, or overall. Patients and Methods: The study cohort included 2226 patients randomized to immediate repair or surveillance for the UK Small Aneurysm Trial (September 1, 1991, through July 31, 1998; follow-up, 2.6-6.9 years) or the Aneurysm Detection and Management trial (August 1, 1992, through July 31, 2000; follow-up, 3.5-8.0 years). Survival differences were assessed with proportional hazard models, adjusted for a comprehensive array of clinical and nonclinical risk factors. Interaction between treatment and AAA size was added to the model to assess whether the effect of immediate open repair vs surveillance varied by AAA size. Results: The adjusted analysis revealed no statistically significant survival difference between immediate open repair and surveillance patients (hazard ratio [HR], 0.99; 95% CI, 0.83-1.18; mean follow-up time, 1921 days for both study groups). This lack of treatment effect persisted when men (HR, 1.01; 95% CI, 0.84-1.21) and women (HR, 0.96; 95% CI, 0.49-1.86) were examined separately and did not vary by AAA size (P¼.39 for the entire cohort and P¼.24 for women). Conclusion: Immediate open repair offered no significant survival benefit, even in patients with the largest AAAs and highest risk of rupture. Because recent trials failed to find a survival benefit of immediate endovascular repair over surveillance for small asymptomatic AAAs, our findings suggest that the gray area of first-line management for these patients should be resolved in favor of surveillance. ª 2013 Mayo Foundation for Medical Education and Research n Mayo Clin Proc. 2013;88(9):910-919 For editorial comment, see page 905 From the Institute for Health Care Research and Improvement, Baylor Health Care System (G.F., D.J.B., B.d.G., J. Harbor, J.B.V.), and Baylor University Medical Center (G.F.), Dallas, TX; Department of Infectious Diseases, University of Louisville, Louisville, KY (G.F.); The Heart Hospital at Baylor Plano, Plano, TX (G.F.); Department of Medicine, Veterans Affairs Affiliations continued at the end of this article. Management of asymptomatic abdominal aortic aneurysms (AAAs) depends on their size. is performed on larger aneurysms (>6.0-cm diameter), whereas very small aneurysms (<4.0-cm diameter), in which the risk of rupture is low, are monitored for growth through regular ultrasonography. 1 For small AAAs (4.0- to 5.4-cm diameter) the risk of rupture is uncertain, and there has been considerable debate regarding the most beneficial course of management: immediate repair vs surveillance and selective repair of AAAs that subsequently enlarge. 2 The 1991 RAND panel assessing appropriateness and necessity of surgery for AAAs found reports of risk of rupture, on the basis of referral case series, as high as 5% per year for AAAs larger than 5.0 cm and 3% to 5% per year for AAAs 5.0 cm or smaller, supporting arguments in favor of the aggressive approach of immediate repair. 3 However, population data suggest that risk of rupture is less than 1% per year, 3,4 under which scenario the merits of surveillance are apparent. Only 2 randomized controlled trials have compared immediate open repair and surveillance for small AAAs: the UK Small Aneurysm Trial (UKSAT) 5-7 and the Aneurysm Detection and Management (ADAM) trial in the United States. 8 Both trials concluded that among patients with small AAAs (4.0-5.4 cm) immediate surgery was not significantly associated with improved long-term survival (maximum followup, >7 years) and carried risks of adverse outcomes, such as impotence. 9 Although it was initially thought that endovascular aneurysm repair (EVAR), with its lower perioperative risks, 910 Mayo Clin Proc. n September 2013;88(9):910-919 n http://dx.doi.org/10.1016/j.mayocp.2013.05.014 n ª 2013 Mayo Foundation for Medical Education and Research

SMALL AAA MANAGEMENT might tip the balance in favor of early intervention for patients with AAAs of 4.0- to 5.4-cm diameter, 10 neither of the 2 randomized controlled trials comparing EVAR and surveillance for AAAs with a diameter less than 5.5 cm (Positive Impact of Endovascular Options for Treating Aneurysms Early [PIVOTAL] trial 11,12 and Comparison of versus Aortic Endografting for Small Aneurysm Repair [CAESAR] trial 13,14 ) reported a survival benefit with early EVAR over surveillance (maximum follow-up, 2 and 4 years), despite much lower perioperative mortality than was seen in the UKSAT and ADAM trial. 14,15 Thus, current guidelines for treatment of patients with AAAs still note, Debate remains for patients presenting with AAAs between 4.0 cm and 5.4 cm regarding the most appropriate role for either immediate treatment or surveillance and selective repair for those aneurysms that subsequently enlarge beyond 5.4 cm. 16 This is particularly true for women because research indicates they are at increased risk of rupture 5,17 and experience greater perioperative and postoperative mortality and morbidity. 5,18-22 Accordingly, we analyzed the combined, original UKSAT and ADAM trial data sets; the 2 smaller, industry-funded trials evaluating EVAR were not included. PATIENTS AND METHODS We compared survival between patients with small AAAs receiving immediate open repair vs surveillance, overall and by sex, using the original patient data from the UKSAT and ADAM trial. We then investigated whether survival differences existed between immediate open repair and surveillance on the basis of AAA size, looking at both the overall combined study population from the UKSAT and ADAM trial and at the subgroup of women included in these 2 trials. Intervention In both trials, patients who were randomly assigned to the immediate open repair group received standard open repair within 6 weeks after randomization. 23,24 Likewise, in both trials patients assigned to surveillance were followed up without repair at similar regular intervals (a minimum of once every 6 months), and surgery was performed within 6 weeks if (1) the aneurysm reached 5.5 cm, (2) the aneurysm enlarged by a minimum of 0.7 cm in 6 months or 1.0 cm in 1 year, or (3) the aneurysm became symptomatic. 23,24 Randomization In both the ADAM trial and the UKSAT, randomization was designed with equal probability of assignment to each of the 2 groups and stratified according to medical center. 5,8 Patients Study Population. The study population consisted of all men and women identified as having an AAA of 4.0 to 5.4 cm in diameter who were randomized to either immediate open surgery or a period of ultrasonographic surveillance and selective surgery in either the ADAM (August 1, 1992, through July 31, 2000) or UKSAT (September 1, 1991, through July 31, 1998) studies. The total number of patients included in the combined study cohort was 2226 (Table). We excluded 88 patients from the adjusted analysis because data on one or more of the covariates listed in the Table below were missing. The final study cohort, therefore, included 2138 patients. Inclusion and Exclusion Criteria. Inclusion and exclusion criteria for the ADAM trial and the UKSAT were comparable. Both trials enrolled patients with small (4.0-5.4 cm), nontender, asymptomatic AAAs distal to the renal arteries considered fit for immediate open repair. 23,24 Both trials excluded patients who were considered unfit for immediate open repair, had symptoms associated with the aneurysm, were unable to attend the follow-up visit, or were unable to give informed consent. 23,24 The ADAM study also excluded patients who underwent a revascularization procedure within 3 months before enrollment, had a myocardial infarction within 6 months before enrollment, or were expected to survive less than 5 years because of invasive cancer or other life-threatening disease. 18 Lastly, age inclusion criteria were 50 to 79 years and 60 to 76 years for the ADAM trial and the UKSAT, respectively. 23,24 Despite the relatively wider age range of patients eligible Mayo Clin Proc. n September 2013;88(9):910-919 n http://dx.doi.org/10.1016/j.mayocp.2013.05.014 911

MAYO CLINIC PROCEEDINGS TABLE. Characteristics of the Study Cohort of 2226 Patients Randomized to Either Immediate Open Repair or in the ADAM Trial and UKSAT a,b (n¼1094) Immediate open repair (n¼1132) Characteristic Median Mean IQR No. (%) Median Mean IQR No. (%) Study ADAM... 567 (51.8)... 569 (50.3) UKSAT... 527 (48.2)... 563 (49.7) Race c White (n¼1068)... 530 (93.5)... 538 (94.6) Black (n¼38)... 25 (4.4)... 13 (2.3) Other/unknown (n¼30)... 12 (2.1)... 18 (3.2) Sex Male... 999 (91.3)... 1030 (91.0) Female... 95 (8.7)... 102 (9.0) Age (y) 69.0 68.5 64.8-72.6. 69.5 68.8 65.0-73.0. Height (cm) 175.0 174.1 170.0-180.3. 175.0 173.9 170.0-180.3. Weight (kg) 79.2 80.4 69.9-89.8. 79.4 80.0 69.9-88.9. BSA (Dubois method) (m 2 ) 1.95 1.95 1.8-2.1. 1.94 1.94 1.8-2.1. Time in the study (d) 1582 1620 1221-2092. 1650 1621 1241-2118. Family history of AAA... 97 (8.9)... 110 (9.7) Smoking Never... 63 (5.8)... 63 (5.6) Previous... 650 (59.43)... 623 (55.0) Current... 381 (34.8)... 446 (39.4) Hypertension... 520 (47.4)... 539 (47.6) Chronic lung disease... 174 (15.9)... 181 (16.0) Stroke... 83 (7.6)... 92 (8.1) Diabetes mellitus... 72 (6.6)... 69 (6.1) Embolism... 14 (1.3)... 9 (0.8) Ischemic heart disease... 442 (40.4)... 466 (41.2) History of MI... 257 (23.5)... 263 (23.2) History of cancer... 127 (11.6)... 124 (11.0) DVT... 36 (3.3)... 32 (2.8) AAA diameter (cm) 4.6 4.7 4.3-5.0. 4.6 4.7 4.3-5.0. Cholesterol (mg/dl) Total 223.5 225.0 194.0-251.0. 222.0 224.6 193.1-251.0. HDL 39.0 42.1 32.8-48.0. 38.6 41.3 32.8-47.0. LDL 144.0 146.8 120.0-170.0. 144.0 146.3 118.3-172.0. Creatinine (mg/dl) 1.1 1.2 1.0-1.3. 1.1 1.2 1.0-1.3. Forced expiratory volume in 1 second (L) 2.4 2.4 1.8-2.8. 2.4 2.4 1.9-2.8. Previous CABG or PCI... 169 (15.5)... 170 (15.0) Medication Aspirin... 428 (39.1)... 419 (37.0) b-blocker... 166 (15.2)... 179 (15.8) Antiarrhythmic... 73 (6.7)... 72 (6.4) Cholesterol lowering... 132 (12.1)... 137 (12.1) Hypertension lowering... 539 (49.3)... 578 (51.1) AAA repair... 678 (61.8)... 1047 (92.5) AAA rupture... 31 (2.8)... 8 (0.7) Repair due to AAA rupture... 14 (1.3)... 5 (0.4) Time to rupture (d) 1579 1617 1221-2097. 1650 1619 1235-2118. Time until surgery, if had surgery (d) 722 808 384-1148. 37 86 26-65. Survival status Alive... 821 (75.0)... 830 (73.3) Dead... 273 (25.0)... 302 (26.7) Continued on next page 912 Mayo Clin Proc. n September 2013;88(9):910-919 n http://dx.doi.org/10.1016/j.mayocp.2013.05.014

SMALL AAA MANAGEMENT TABLE. Continued (n¼1094) Immediate open repair (n¼1132) Characteristic Median Mean IQR No. (%) Median Mean IQR No. (%) Survival status, continued AAA-related death for total population... 54 (4.9)... 51 (4.5) Death within 30 d of surgery... 29 (2.7, d 4.3 e )... 40 (3.5, d 3.8 e ) Time to death or censoring (d surviving). 1621... 1621.. a AAA ¼ abdominal aortic aneurysm; ADAM ¼ Aneurysm Detection and Management; BSA ¼ body surface area; CABG ¼ coronary artery bypass graft ; DVT ¼ deep venous thrombosis; HDL ¼ high-density lipoprotein; IQR ¼ interquartile range; LDL ¼ low-density lipoprotein; MI ¼ myocardial infarction; PCI ¼ percutaneous coronary intervention; UKSAT ¼ UK Small Aneurysm Trial. b SI conversion factors: To convert cholesterol (total, HDL, and LDL) to mmol/l, multiply by 0.0259; to convert creatinine to mmol/l, multiply by 88.4. c Race data were not collected for the UKSAT. d Percentage of total population (n surveillance ¼ 1094; n immediate open repair ¼ 1132). e Percentage of the population that underwent AAA repair (n surveillance ¼ 678; n immediate open repair ¼ 1047). for inclusion in the ADAM trial, most participants (88%) fell within the same age range as those in the UKSAT. More than 92% of the combined study cohort was aged 60 to 76 years. Subgroups of Interest. We investigated whether survival differences existed between immediate open repair and surveillance, overall and within the subgroups of men and women. We also investigated survival differences by treatment assignment depending on AAA diameter, overall and within the subgroup of women, because of the ongoing debate about whether the size threshold for surgery should be lowered for women. Study Measures Outcome of Interest. ADAM trial vital status was assessed using direct contact, Department of Veterans Affairs databases, the Social Security Administration s Death Master File, and the National Death Index, and was defined as days from randomization to either death or July 31, 2000. Vital status for the UKSAT was assessed using the Office for National Statistics notifications of patient deaths and was defined as days from randomization to either death or July 31, 1998. Follow-up ranged from 2.6 to 6.9 years in the UKSAT and from 3.5 to 8.0 years in the ADAM trial. Independent Variables of Interest. We considered 2 main independent variables: treatment, defined as intervention assignment (to either immediate open repair or surveillance) at the time of randomization in the UKSAT and ADAM trial, and AAA size. Covariates. Several potential confounders of the associations of interest were considered: (1) a number of baseline factors, including age, sex, diabetes mellitus, smoking, hypertension, evidence of myocardial infarction, creatinine level, total cholesterol level, forced expiratory volume in 1 second, aspirin use, b- blocker use, antiarrhythmic medication use, and cholesterol-lowering medication use, in addition to treatment group and trial effect (ie, ADAM trial or the UKSAT) and (2) a number of other baseline factors (using the requested variables), including body surface area defined as (0.007184) (mass in kilograms ^ 0.425) (height in centimeters ^ 0.725); previous revascularization, defined as the patient having previously undergone coronary artery bypass graft or percutaneous coronary intervention; history of ischemic disease, defined as the patient having a history of myocardial infarction, angina or coronary artery disease, or stroke or transient stroke; history of pulmonary disease, defined as the patient having emphysema, chronic obstructive pulmonary disease, or pulmonary emboli; and history of cancer. Statistical Analyses The Kaplan-Meier method was used to construct a survival curve for each trial arm (immediate repair vs surveillance). A Cox proportional hazards model was used to investigate whether there was a survival difference for immediate open repair vs surveillance in the combined Mayo Clin Proc. n September 2013;88(9):910-919 n http://dx.doi.org/10.1016/j.mayocp.2013.05.014 913

MAYO CLINIC PROCEEDINGS ADAM trial and UKSAT cohort of patients; the dependent variable was time from randomization until death or last follow-up. This model included the trial arm as the sole independent variable. A Wald statistic was calculated from this model to investigate the difference between trial arms; a hazard ratio (HR) and accompanying 95% CIs were also calculated. To account for any residual randomization imbalance, a generalized propensity approach was used to estimate the adjusted difference in survival between the immediate open repair and surveillance groups. The propensity score was created by modeling trial arm as a function of study (UKSAT vs ADAM trial), AAA diameter, age, body surface area, sex, smoking status, forced expiratory volume in 1 second, chronic obstructive pulmonary disease, diabetes, hypertension, previous revascularization, creatinine level, total cholesterol level, history of cancer, history of myocardial infarction, ischemic heart disease, and the use of the baseline medications listed in the Table. The propensity score was modeled with a cubic spline with 3 knots 25 and was included in the Cox model along with the intervention (immediate open repair vs surveillance). The unadjusted and adjusted analyses were repeated for men and women separately. An additional Cox survival model was fitted, including the trial arm, AAA size, and an interaction between trial arm and AAA size as independent variables to investigate whether the effect of trial arm varied by AAA size. Again, to avoid problems regarding the assumption of a linear effect of AAA size on survival, AAA size was modeled using a restricted cubic spline with 3 knots. 25,26 Because the patients were obviously not randomized to AAA diameter, the risk factors listed in the Table could not be assumed to be balanced across AAA size; thus, the Cox model was adjusted for the same covariates listed for the previous model via a generalized propensity approach. 27 Briefly, a linear regression model was fitted with AAA size as a continuous dependent variable and with each covariate as an independent variable. Letting y (1) represent the predicted value (Xb) from this propensity model, AAA (1) denote patient AAA diameter, and ARM denote the trial arm (0 for surveillance and 1 for immediate repair), the final adjusted Cox model investigating the interaction between trial arm and AAA diameter was as follows: logðhðtþþ ¼ g 0 þ g 1 ARM þ g 2 AAA ð1þ þ g 3 AAA ð2þ þ g 4 ARM, Here, AAA ð1þ þ AAA ð2þ þ g 5 y ð1þ þ g 6 y ð2þ y ð2þ ¼ a þ 1 aþ 2 ðk 3 k 1 Þ=ðk 3 k 2 Þ þ a þ 3 ðk 2 k 1 Þ=ðk 3 k 2 Þ in which k j denotes the j th knot for the cubic spline and a þ j ¼ðy ð1þ k j Þ 3 if y ð1þ > k j and ¼ 0 otherwise. The terms AAA (1) and AAA (2) form similar component variables for the cubic spline fit of AAA size. The P value for the adjusted effect of immediate open repair vs surveillance was determined on the basis of a Wald statistic for the interaction term from this model. The probability of survival at 1 and 5 years vs AAA diameter was also plotted for each trial arm on the basis of this model. The same analysis of the interaction between AAA diameter and treatment described previously was executed for the subgroup of women. The test statistic of Grambsch and Therneau 28 was used to test for departures from the proportional hazards assumption for each of the factors included in the previously described Cox models. Interaction between the study (ADAM trial or the UKSAT) effect and intervention (immediate open repair or surveillance) was also investigated but was not statistically significant at the a¼.05 level and therefore was not included in the final models. All analyses were performed with the R software, version 2.14.1. 29 An a of.05 was used for all analyses to determine statistical significance. RESULTS The Table lists the characteristics of the study population by treatment group. The ADAM trial and the UKSAT were of similar size, so neither study substantially outweighed the other in the overall study population (approximately 53% of patients were participants in the ADAM trial compared with 47% in the UKSAT). In the overall study population, more than 90% of patients in both treatment groups were white and male, the demographic group known to have the highest incidence of AAA. 30 Median patient age, weight, height, body surface area, and number 914 Mayo Clin Proc. n September 2013;88(9):910-919 n http://dx.doi.org/10.1016/j.mayocp.2013.05.014

SMALL AAA MANAGEMENT of days in the study were similar between the treatment groups, as was the prevalence of all clinical risk factors and preoperative use of medications. The overall unadjusted survival curves for the immediate open repair and surveillance groups are presented in Figure 1. No statistically significant effect of treatment (immediate open repair vs surveillance) was observed in the overall study population (HR, 0.96; 95% CI, 0.81-1.14; P¼.64) or in either the subgroup of men (HR, 0.96; 95% CI, 0.80-1.15; P¼.62) or women (HR, 0.96; 95% CI, 0.52-1.77; P¼.90). Likewise, the adjusted Cox analysis revealed no survival differences between treatment groups in the overall study population (HR, 0.99; 95% CI, 0.83-1.18), the cohort of men (HR, 1.01; 95% CI, 0.84-1.21), or the cohort of women (HR, 0.96; 95% CI, 0.49-1.86). Moreover, the interaction between AAA size and treatment assignment was not significant in the same adjusted Cox model (P¼.39). The propensity-adjusted probabilities of survival by AAA diameter and treatment at 1 and 5 years on the basis of this interaction model are presented in Figure 2. These results indicate no statistically significant difference in survival between immediate open repair and surveillance, regardless of AAA diameter. Small differences in the anticipated direction (surveillance more favorable for smaller sizes and immediate repair for larger sizes) were observed at 5 years, but the large P value of.39 precludes any conclusion regarding superiority of either method that should influence clinical decision making. The interaction of AAA diameter and treatment in the subgroup of women was also nonsignificant (P¼.24), indicating that the lack of survival difference between interventions did not vary by AAA diameter in the 179 women in the study cohort who were not missing data for relevant variables (18 women [9.1%] had missing data). The propensityadjusted probabilities of survival by AAA diameter and treatment at 1 and 5 years are presented in Figure 3. DISCUSSION Given the nearly identical design of the UKSAT and ADAM trial, our analysis of the combined detailed, original clinical data from these studies essentially amounts to a large, multicenter randomized controlled trial comparing immediate open repair and surveillance for small asymptomatic AAAs. We found no difference in overall survival regardless of AAA size within the 4.0- to 5.4-cm range. We also found no differences in survival in the subgroups of either men or women. These results support the evidence underlying the Society for Vascular s strong recommendation in favor of surveillance for patients with a fusiform AAA of 4.0 to 5.4 cm. 31 Our results also extend the findings of the recent Cochrane Systematic Review (which concluded that immediate open repair for small asymptomatic AAAs is not supported by current evidence) 32 because no survival benefits were observed with immediate open surgery, even in patients expected to have higher risk of rupture because of relatively larger AAAs. Moreover, although the UKSAT and ADAM trial considered only open surgery for AAA repair, the failure of both the PIVOTAL and CAESAR trials to find any survival benefit of immediate EVAR over surveillance at 20 and 32 months of follow-up, respectively, 14,15 Cumulative probability of death 0.4 0.3 0.2 0.1 No. at risk 0 0 1 2 3 4 5 6 7 Follow-up (y) 1077 1018 980 926 667 441 235 67 1061 1027 980 912 615 397 231 76 FIGURE 1. Kaplan-Meier curves depicting the cumulative probability of death by treatment for 2138 patients randomized to either immediate open repair or surveillance in the Aneurysm Detection and Management trial or the United Kingdom Small Aneurysm Trial who had complete follow-up (hazard ratio, 0.96; 95% CI, 0.81-1.14; P¼.64). Mayo Clin Proc. n September 2013;88(9):910-919 n http://dx.doi.org/10.1016/j.mayocp.2013.05.014 915

MAYO CLINIC PROCEEDINGS 1-Year survival probability (%) 5-Year survival probability (%) 100 80 60 40 100 80 60 40 4.0 4.5 5.0 5.5 AAA diameter (cm) 4.0 4.5 5.0 5.5 AAA diameter (cm) FIGURE 2. Adjusted (Cox) 1-year and 5-year survival probability by treatment and by abdominal aortic aneurysm (AAA) diameter for 2138 patients randomized to either immediate open repair or selective surveillance in the Aneurysm Detection and Management (ADAM) trial or the United Kingdom Small Aneurysm Trial (UKSAT) who had complete followup. Propensity-adjusted P¼.39 (adjusted for study [UKSAT vs ADAM], age, body surface area, sex, smoking status, forced expiratory volume in 1 second, chronic obstructive pulmonary disease, diabetes mellitus, hypertension, previous revascularization, creatinine level, total cholesterol level, history of cancer, history of myocardial infarction, ischemic heart disease, and the use of aspirin, b-blockers, antiarrhythmic medications, or cholesterol-lowering medications). suggests that the gray area of first-line management of asymptomatic small AAAs can be resolved in favor of surveillance and that resources should be focused on ensuring patients have access to and adhere to effective surveillance programs. Some surgeons have suggested that the recommended threshold for immediate repair be decreased to 5.0 cm for women 33 (on the basis of the increased risk of rupture of small AAAs previously reported in women compared with men 5,17 ), whereas others point out that this risk may be outweighed by the greater perioperative and postoperative mortality and morbidity experienced by women. 5,18-21 The subgroup analysis we conducted in the women from the UKSAT and ADAM trial attempted to address this question and found no association between survival differences with immediate repair vs surveillance and AAA diameter. Although there appear to be differences in 5- year survival (favoring surveillance) for some AAA diameters as shown in Figure 3, the 95% CIs of the estimated probability function are very broad because of the limited number of women in our study population. All our results for the subgroup of women must be interpreted with extreme caution because there were only 179 women in the combined UKSAT and ADAM trial populations. A larger study is needed to reach a definitive answer on the appropriate threshold for repair in women. A more general call to lower the threshold for immediate AAA repair has also been made on the basis of the lower risk of operative morality with EVAR. 34 However, the combination of the lack of survival benefit seen in the intermediate results of the PIVOTAL and CAESAR trials, the extrapolation of our results from the long-term follow-up of the ADAM trial and the UKSAT indicating that the lack of survival benefit with immediate repair is consistent across diameter size within small AAAs, and the higher costs of EVAR and the associated need for reintervention and longterm follow-up surveillance found in some randomized controlled trials of repair for larger AAAs and observational studies 35-38 (although the recent cost-effectiveness comparison of EVAR vs open repair in the Veterans Affairs Open Versus Endovascular Repair trial reported nonsignificantly lower costs in the EVAR group at 2 years 39 ) may suggest that use of EVAR for small AAAs should decrease rather than increase and that surveillance should become the evidence-based first-line management strategy. Under this scenario, research is needed to investigate whether open repair or EVAR is more effective (in terms of long-term survival, durability, and quality of life) when AAA expansion or development of symptoms indicates repair is needed. The possibility of medical management of AAAs 916 Mayo Clin Proc. n September 2013;88(9):910-919 n http://dx.doi.org/10.1016/j.mayocp.2013.05.014

SMALL AAA MANAGEMENT has also recently entered the debate, 40 and 4 randomized controlled trials are currently examining the efficacy of different medical management strategies. 41 However, in the absence of a proven treatment, repair remains the dominant question for clinical practice. Limitations of this study include the limitations that applied to the original UKSAT and ADAM trial from which we analyzed the combined data. Important among these was the substantial underrepresentation of minorities (not reported in the UKSAT and 6% in the ADAM trial). 8,42 As such, our results may not be generalizable to these subgroups of patients and not add substantially to the evidence already available from the ADAM trial and the UKSAT on differences that may exist among races. In addition, our results are based on trials conducted more than 10 years ago. Although this is necessary to obtain the longterm survival data we were interested in, and necessitated by the fact there have been no further trials comparing surveillance and open repair for small AAAs, changes in operative technique and perioperative care, as well as in accuracy of surveillance imaging technology and medical management (aneurysm related or more generally for the increased risk of cardiovascular events and atherosclerosis in AAA patients 40 ), might have altered in the interim. Our results stem from randomized controlled trial settings that, particularly for the surveillance group, may not reflect either the resources available for care or patient adherence to follow-up schedules that can be expected in regular practice settings. As such, we conclude that there was no significant difference in efficacy between open repair and surveillance in small AAAs; every effort should be made to ensure the availability of resources for and patient adherence to adequate and rigorous follow-up surveillance. Although the long-term results of the PIVOTAL and CAESAR trials remain of interest, future research needs to shift focus from the procedure as a solution mind-set that has dominated AAA research and management in the past and instead focus on what is still unknown about the disease itself. Most urgently, a large, prospective, population-based study is needed to assess the progression of disease in relation to AAA morphologic 1-Year survival probability (%) 5-Year survival probability (%) 100 80 60 40 4.0 4.5 5.0 5.5 AAA diameter (cm) 100 80 60 40 4.0 4.5 5.0 5.5 AAA diameter (cm) FIGURE 3. Adjusted (Cox) 1-year and 5-year survival probability by treatment and by abdominal aortic aneurysm (AAA) diameter for 179 women randomized to either immediate open repair or selective surveillance in the Aneurysm Detection and Management (ADAM) trial or the United Kingdom Small Aneurysm Trial (UKSAT) who had complete follow-up. Propensityadjusted P¼.24 (adjusted for study [UKSAT vs ADAM], age, body surface area, sex, smoking status, forced expiratory volume in 1 second, chronic obstructive pulmonary disease, diabetes mellitus, hypertension, previous revascularization, creatinine level, total cholesterol level, history of cancer, history of myocardial infarction, ischemic heart disease, and the use of aspirin, b-blockers, antiarrhythmic medications, or cholesterol-lowering medications). features (including shape, size, location, volume, and ratio of healthy aorta to the aneurysm), particularly looking at whether AAA volume is superior to diameter as a measure of progression of AAA disease. Research is also needed to determine whether efficacy or effectiveness of the various treatment options (open repair, EVAR, and the emerging medical management options) differs on the basis of AAA morphologic features. Such detailed evaluation would not only inform the development of new, more targeted treatment regimens and associated evidence-based guidelines but also, Mayo Clin Proc. n September 2013;88(9):910-919 n http://dx.doi.org/10.1016/j.mayocp.2013.05.014 917

MAYO CLINIC PROCEEDINGS at the level of the individual physician-patient interaction, inform the shared decision-making process and provide both the patient and the clinician with greater confidence that they have selected the best option for that patient s needs. CONCLUSION Immediate open repair offered no significant survival benefit, overall or in either of the subgroupsofmenandwomen,eveninpatients with the highest risk of rupture (ie, the largest AAAs). Combined with the failure of recent trials to find a survival benefit ofimmediate EVAR over surveillance for small asymptomatic AAAs, our findings suggest that the gray area of first-line management for these patients should be resolved in favor of surveillance. Abbreviations and Acronyms: AAA = abdominal aortic aneurysm; ADAM = Aneurysm Detection and Management; CAESAR = Comparison of versus Aortic Endografting for Small Aneurysm Repair; EVAR = endovascular aneurysm repair; HR = hazard ratio; PIVOTAL = Positive Impact of Endovascular Options for Treating Aneurysms Early; UKSAT = United Kingdom Small Aneurysm Trial Affiliations (Continued from the first page of this article.): Medical Center, Minneapolis, MN (F.A.L.); Department of Biostatistics, Heart Valve Therapies, Edwards Lifesciences, Irvine, CA (C.H.); Department of Cardiology, Yale University School of Medicine, New Haven, CT (J. Herrin); Health Research and Educational Trust, Chicago, IL (J. Herrin); VA Cooperative Studies Program Coordinating Center, West Haven, CT (G.R.J.); and Vascular, Imperial College at Charing Cross, London, UK (J.T.P.). Grant Support: The ADAM trial was supported by the Veterans Affairs Cooperative Studies Program, Department of Veterans Affairs Office of Research and Development, Washington, DC. The UKSAT was funded by the Medical Research Council and British Heart Foundation. This work was funded by Agency for Healthcare Research and Quality grant R01HS018576. Additional funding was provided by the Bradley Family Endowment to Baylor Health Care System Foundation. Potential Competing Interests: The authors acknowledge the use of software from Professor Frank Harrell s Hmisc and rms libraries. Dr Hamilton discloses an employment relationship with Edwards Lifesciences Inc. Correspondence: Address to Giovanni Filardo, PhD, MPH, Institute for Health Care Research and Improvement, 8080 North Central Expressway, Ste 900, Dallas, TX 76206 (giovanfi@baylorhealth.edu). REFERENCES 1. Hirsch AT, Haskal ZJ, Hertzer NR, et al. ACC/AHA Guidelines for the Management of Patients with Peripheral Arterial Disease (lower extremity, renal, mesenteric, and abdominal aortic): a collaborative report from the American Associations for Vascular /Society for Vascular, Society for Cardiovascular Angiography and Interventions, Society for Vascular Medicine and Biology, Society of Interventional Radiology, and the ACC/AHA Task Force on Practice Guidelines (writing committee to develop guidelines for the management of patients with peripheral arterial disease)esummary of recommendations. J Vasc Interv Radiol. 2006;17(9):1383-1398. 2. Lederle FA. Risk of rupture of large abdominal aortic aneurysms: disagreement among vascular surgeons. Arch Intern Med. 1996;156(9):1007-1009. 3. Ballard DJ, Etachason JA, Hilborne LH, et al. Abdominal Aortic Aneurysm : A Literature Review and Ratings of Appropriateness and Necessity. Santa Monica, CA: RAND; 1992. 4. Nevitt MP, Ballard DJ, Hallett JW Jr. Prognosis of abdominal aortic aneurysms: a population-based study. N Engl J Med. 1989;321(15):1009-1014. 5. United Kingdom Small Aneurysm Trial Participants. Long-term outcomes of immediate repair compared with surveillance of small abdominal aortic aneurysms. N Engl J Med. 2002; 346(19):1445-1452. 6. Mortality results for randomised controlled trial of early elective surgery or ultrasonographic surveillance for small abdominal aortic aneurysms: The UK Small Aneurysm Trial Participants. Lancet. 1998;352(9141):1649-1655. 7. Powell JT. Final 12-year follow-up of surgery versus surveillance in the UK Small Aneurysm Trial: UK Small Aneurysm Trial participants. Br J Surg. 2007;94(6):702-708. 8. Lederle FA, Wilson SE, Johnson GR, et al. Immediate repair compared with surveillance of small abdominal aortic aneurysms. N Engl J Med. 2002;346(19):1437-1444. 9. Lederle FA, Johnson GR, Wilson SE, et al. Quality of life, impotence, and activity level in a randomized trial of immediate repair versus surveillance of small abdominal aortic aneurysm. J Vasc Surg. 2003;38(4):745-752. 10. Zarins CK, Crabtree T, Bloch DA, Arko FR, Ouriel K, White RA. Endovascular aneurysm repair at 5 years: does aneurysm diameter predict outcome? J Vasc Surg. 2006;44(5):920-931. 11. Ouriel K. The PIVOTAL study: a randomized comparison of endovascular repair versus surveillance in patients with smaller abdominal aortic aneurysms. J Vasc Surg. 2009;49(1):266-269. 12. King C, Papillon S. Medtronic and the Cleveland Clinic launch clinical trial evaluating early treatment of small aneurysms. 2005. http://wwwp.medtronic.com/newsroom/newsreleasedetails. do?itemid¼1120751150202&lang¼en_us. Accessed October 31, 2006. 13. Cao P. Comparison of surveillance vs Aortic Endografting for Small Aneurysm Repair (CAESAR) trial: study design and progress. Eur J Vasc Endovasc Surg. 2005;30(3):245-251. 14. Cao P, De Rango P, Verzini F, Parlani G, Romano L, Cieri E. Comparison of surveillance versus aortic endografting for small aneurysm repair (CAESAR): results from a randomised trial. Eur J Vasc Endovasc Surg. 2011;41(1):13-25. 15. Ouriel K, Clair DG, Kent KC, Zarins CK. Endovascular repair compared with surveillance for patients with small abdominal aortic aneurysms. J Vasc Surg. 2010;51(5):1081-1087. 16. Chaikof EL, Brewster DC, Dalman RL, et al. The care of patients with an abdominal aortic aneurysm: the Society for Vascular practice guidelines. J Vasc Surg. 2009;50(4, suppl):s2-s49. 17. Brown PM, Zelt DT, Sobolev B. The risk of rupture in untreated aneurysms: the impact of size, gender, and expansion rate. J Vasc Surg. 2003;37(2):280-284. 18. Abedi NN, Davenport DL, Xenos E, Sorial E, Minion DJ, Endean ED. Gender and 30-day outcome in patients 918 Mayo Clin Proc. n September 2013;88(9):910-919 n http://dx.doi.org/10.1016/j.mayocp.2013.05.014

SMALL AAA MANAGEMENT undergoing endovascular aneurysm repair (EVAR): an analysis using the ACS NSQIP dataset. J Vasc Surg. 2009;50(3):486-491:e481-e484. 19. Mehta M, Byrne WJ, Robinson H, et al. Women derive less benefit from elective endovascular aneurysm repair than men. J Vasc Surg. 2012;55(4):906-913. 20. Wisniowski B, Barnes M, Jenkins J, Boyne N, Kruger A, Walker PJ. Predictors of outcome after elective endovascular abdominal aortic aneurysm repair and external validation of a risk prediction model. J Vasc Surg. 2011;54(3):644-653. 21. McPhee JT, Hill JS, Eslami MH. The impact of gender on presentation, therapy, and mortality of abdominal aortic aneurysm in the United States, 2001-2004. J Vasc Surg. 2007;45(5):891-899. 22. Ballard DJ, Filardo G, Fowkes G, Powell JT. for small asymptomatic abdominal aortic aneurysms. Cochrane Database Syst Rev. 2008;4:CD001835. 23. The U.K. Small Aneurysm Trial: design, methods and progress. The UK Small Aneurysm Trial participants. Eur J Vasc Endovasc Surg. 1995;9(1):42-48. 24. Lederle FA, Wilson SE, Johnson GR, et al; ADAM VA Cooperative Study Group. Design of the abdominal aortic Aneurysm Detection and Management Study. J Vasc Surg. 1994;20(2): 296-303. 25. Harrell FE Jr. Regression Modeling Strategies: With Application to Linear Models, Logistic Regression, and Survival Analysis. New York, NY: Springer-Verlag; 2001. 26. Filardo G, Hamilton C, Hamman B, Ng HK, Grayburn P. Categorizing BMI may lead to biased results in studies investigating in-hospital mortality after isolated CABG. J Clin Epidemiol. 2007;60(11):1132-1139. 27. Imai K, van Dyk DA. Causal inference with general treatment regimes: generalizing the propensity score. J Am Stat Assoc. 2004;99(467):854-866. 28. Grambsch PM, Therneau T. Proportional hazards tests and diagnostics based on weighted residuals. Biometrika. 1994;81(3): 515-526. 29. R version 2.7.1 [computer program]. www.r-project.org. Accessed June 26, 2013. 30. Wilson CT, Fisher E, Welch HG. Racial disparities in abdominal aortic aneurysm repair among male Medicare beneficiaries. Arch Surg. 2008;143(5):506-510. 31. Chaikof EL, Brewster DC, Dalman RL, et al. SVS practice guidelines for the care of patients with an abdominal aortic aneurysm: executive summary. J Vasc Surg. 2009; 50(4):880-896. 32. Filardo G, Powell JT, Martinez MA, Ballard DJ. for small asymptomatic abdominal aortic aneurysms. Cochrane Database Syst Rev. 2012;3:CD001835. 33. Brewster DC, Cronenwett JL, Hallett JW Jr, Johnston KW, Krupski WC, Matsumura JS. Guidelines for the treatment of abdominal aortic aneurysms. Report of a subcommittee of the Joint Council of the American Association for Vascular and Society for Vascular. J Vasc Surg. 2003;37(5): 1106-1117. 34. Paraskevas KI, Mikhailidis DP, Veith FJ. The rationale for lowering the size threshold in elective endovascular repair of abdominal aortic aneurysm. J Endovasc Ther. 2011;18(3): 308-313. 35. Epstein DM, Sculpher MJ, Manca A, et al. Modelling the longterm cost-effectiveness of endovascular or open repair for abdominal aortic aneurysm. Br J Surg. 2008;95(2):183-190. 36. Greenhalgh RM, Brown LC, Powell JT, Thompson SG, Epstein D, Sculpher MJ. Endovascular versus open repair of abdominal aortic aneurysm. N Engl J Med. 2010;362(20): 1863-1871. 37. Prinssen M, Buskens E, de Jong SE, et al. Cost-effectiveness of conventional and endovascular repair of abdominal aortic aneurysms: results of a randomized trial. J Vasc Surg. 2007;46(5): 883-890. 38. Tarride JE, Blackhouse G, De Rose G, et al. Should endovascular repair be reimbursed for low risk abdominal aortic aneurysm patients? evidence from Ontario, Canada. Int J Vasc Med. 2011; 2011:308685. 39. Stroupe KT, Lederle FA, Matsumura JS, et al. Cost-effectiveness of open versus endovascular repair of abdominal aortic aneurysm in the OVER trial. J Vasc Surg. 2012;56(4):901-909:e902. 40. Powell JT. Non-operative or medical management of abdominal aortic aneurysm. Scand J Surg. 2008;97(2):121-124. 41. Golledge J, Norman PE. Current status of medical management for abdominal aortic aneurysm. Atherosclerosis. 2011;217(1): 57-63. 42. Mortality results for randomised controlled trial of early elective surgery or ultrasonographic surveillance for small abdominal aortic aneurysms: The United Kingdom Small Aneurysm Trial Participants. Lancet. 1998;352(9141):1649-1655. Mayo Clin Proc. n September 2013;88(9):910-919 n http://dx.doi.org/10.1016/j.mayocp.2013.05.014 919